hushem_5x_deit_tiny_sgd_0001_fold5

This model is a fine-tuned version of facebook/deit-tiny-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 1.4076
• Accuracy: 0.3902

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.0001
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.5266	1.0	28	1.7430	0.2439
1.4534	2.0	56	1.7173	0.2439
1.4589	3.0	84	1.6934	0.2439
1.4589	4.0	112	1.6722	0.2439
1.4564	5.0	140	1.6529	0.2439
1.475	6.0	168	1.6343	0.2439
1.4209	7.0	196	1.6174	0.2439
1.4499	8.0	224	1.6010	0.2439
1.482	9.0	252	1.5863	0.2439
1.4359	10.0	280	1.5742	0.2439
1.4148	11.0	308	1.5620	0.2439
1.4175	12.0	336	1.5508	0.2439
1.4309	13.0	364	1.5414	0.2439
1.4434	14.0	392	1.5314	0.2439
1.4301	15.0	420	1.5227	0.2195
1.4116	16.0	448	1.5140	0.2439
1.4213	17.0	476	1.5063	0.2439
1.4048	18.0	504	1.4983	0.2439
1.4279	19.0	532	1.4909	0.2439
1.422	20.0	560	1.4842	0.2683
1.3928	21.0	588	1.4779	0.2683
1.3839	22.0	616	1.4719	0.2683
1.393	23.0	644	1.4665	0.2683
1.3889	24.0	672	1.4611	0.2683
1.3853	25.0	700	1.4568	0.2927
1.349	26.0	728	1.4521	0.3171
1.3871	27.0	756	1.4479	0.3171
1.3753	28.0	784	1.4439	0.3415
1.3905	29.0	812	1.4406	0.3415
1.3675	30.0	840	1.4371	0.3415
1.3814	31.0	868	1.4338	0.3415
1.3666	32.0	896	1.4305	0.3415
1.3609	33.0	924	1.4275	0.3659
1.3514	34.0	952	1.4249	0.3902
1.3706	35.0	980	1.4224	0.3902
1.3609	36.0	1008	1.4204	0.3902
1.3195	37.0	1036	1.4184	0.3902
1.3869	38.0	1064	1.4165	0.3902
1.3586	39.0	1092	1.4149	0.3902
1.3675	40.0	1120	1.4134	0.3902
1.3299	41.0	1148	1.4121	0.3902
1.3616	42.0	1176	1.4109	0.3902
1.3801	43.0	1204	1.4099	0.3902
1.3528	44.0	1232	1.4091	0.3902
1.3348	45.0	1260	1.4084	0.3902
1.353	46.0	1288	1.4080	0.3902
1.3488	47.0	1316	1.4077	0.3902
1.3749	48.0	1344	1.4076	0.3902
1.3556	49.0	1372	1.4076	0.3902
1.3606	50.0	1400	1.4076	0.3902

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu118
• Datasets 2.15.0
• Tokenizers 0.15.0